The rapid temporal increase in the rates of non-Hodgkin's lymphoma (NHL) suggests there are environmental determinants of the disease, but presently known risk factors can account for only a modest percentage of the increase. A case-control study among women is proposed to investigate several etiologic hypotheses bearing on the increasing rates. Particular emphasis will be placed upon the history of hair-dye use, since four recent studies have suggested an association with NHL but had significant limitations. It is also planned to investigate exposures to pesticides, organic solvents, and certain medications (e.g., anti-inflammatory, immunosuppressive , beta-blockers) as risk factors for these cancers. The rationale for selection of the chemical/drug exposures to investigate is based upon suggestions from prior studies of an NHL association and indications that the agent affects beta-cell or T-cell proliferation or function. This will be a population-based case-control study in upstate New York State among women of ages 20-79 and will include incident cases of NHL found in a three-year period. Because of the high mortality rates for NHL, a fast-reporting system will be instituted in order to interview the cases soon after diagnosis. A histopathological review will be conducted to confirm the case diagnoses and classify the NHL subtypes. it is anticipated that about 600 cases of NHL will be included in the study. For comparison purposes, controls will be selected from the Department of Motor Vehicle files (ages less 65) or from the Health Care Financing Administration (Medicare) files (ages more 65) or from the Health Care Financing Administration (Medicare) files (ages more 65). A telephone interview will be used to obtain information on the risk factors and exposures of interest along with basic sociodemographic information. The hair-dy and other exposure data will be analyzed for total NHL and for several histopathologic subcategories of NHL. Exposure response relationships will be examined with respect to hair dye and other exposures. This study will have good statistical power to test hypotheses about the etiology of non-Hodgkin's lymphoma and to provide relatively precise estimates of relative risk.